Submit matplotlib style to https://github.com/scikit-hep/mplhep

[olantwin]

This would avoid duplication and make it easily available. https://github.com/scikit-hep/mplhep also has some other quality-of-life improvements such as proper histograms and presets for common HEP plots.

For those who don't know it, scikit-hep is a very active community with users and developers from most of the LHC collaborations, focusing mostly on HEP python packages and tooling.

[olantwin]

Idem for ROOT and the ROOT team

[jonas-eschle]

Hi @olantwin I was just going to propose this exact same thing :)

The port can be more: the simulation, preliminary etc text are also available in mplhep under the experiments name (see i.e. the labels for cms

Do you think this would be feasible? I can surely help on the mplhep side (and, for backwards compatibility, mplhep could just be imported here). @chenel you seem to be active in this repo?

Would this be something the DUNE community would be interested in probably? I think it's overall beneficial if the Python HEP communities come closer together

(if maintenance is a concern, Scikit-HEP is open for collaborators in general)

[chenel]

@jonas-eschle sorry I've left this hanging for so long. Unfortunately Oliver has moved on from DUNE, and most of the folks who have assembled this package have moved on to other things. I'm mostly in a "support" rather than "development" phase for this package, myself. So without somebody who is enthusiastic about doing whatever converge work there is available, I don't know how likely this is to go anywhere. Having said that, if anybody who is excited about that arises, I will be sure to point them this way 😄

[jonas-eschle]

No worries, thanks for coming back to this! Yes, I see... I think this is a common issue in HEP with maintenance of packages over time...

I took the liberty therefore to make a PR with the style into mplhep so that the code is streamlined and follows the rest of the other styles of experiments (this is for Python only, not ROOT). As other experiments (CMS, ATLAS, LHCb, ALICE, soon BELLE II) are already there, I think it makes sense to add it there too, so it's the same API and all of that. If you have any comments, feel free to add them! It would also remove the burden a bit from this package, as Scikit-HEP maintains this (an aggregation of physicists/coders, under the umbrella of the HEP software foundation, that provides well maintained packages with tests, docs etc, and people can join if they intend to maintain a specific package).

Does that sound good? Or would you propose a different path maybe? I can also gladly add additional references and acknowledgements in the code-base, or if you're interested to be author of mplhep, just let me know.

[chenel]

I think DUNE folks will probably come and go with updates to our package here as needed. Also, it will be important to us that the C++/PyRoot variant produces the same visual style as the matplotlib one, so we can't push any changes backwards here unless it handles both. So, so long as we can do that freely, I don't think there's any objection to having it linked somewhere else upstream.